For nearly a century, the clicial significance of lumbosacral transitional vertebra (LSTV) has been debated. Simple radiographic findings of 200 patients without low back pain, 200 patients with low back pain and 200 patients(227 discs) with disc abnormalities on CT scan and myelography were reviewed. The prevalence of LSTV of each group according to Castellvis classification, which was based upon radiomorphological change of transverse process of the last presacral vertebra, and the relationship between herniated nucleus pulposus (HNP) and LSTV were investigated. The results were as follows: 1. The prevalence of LSTV using Castellvis classification revealed 59.5% in control group, 54.596 in low backache group, and 57.5% in HNP group, but no significant difference in each group. The prevalence of true LSTV (type II, III, IV) was significantly higher in type II and III in low backache group and type II in HNP group (P < 0.05). 2. In HNP group, the relationship between the type and presence of LSTV and presence of HNP of that level was not different significantly and the true LSTV and the presence of HNP of that level did not have any correlation. 3. The prevalence of HNP at the level above transitional vertebra in LSTV patients was significantly higher than that of herniation in patients who did not have LSTV (P < 0.05).